Provides Emacs font-lock, indentation, navigation and refactoring for the Clojure(Script) programming language.
This document assumes you're familiar with Emacs. More thorough walkthroughs, targeting Emacs beginners, are available at clojure-doc.org and Clojure for the Brave and the True. Keep in mind, however, that they might be out-of-date.
This documentation tracks the master
branch of clojure-mode
. Some of
the features and settings discussed here might not be available in
older releases (including the current stable release). Please, consult
the relevant git tag (e.g. 5.1.0) if you need documentation for a
specific clojure-mode
release.
- Installation
- Bundled major modes
- Configuration
- Refactoring support
- Related packages
- REPL Interaction
- Changelog
- License
Available on the major package.el
community maintained repos -
MELPA Stable and MELPA repos.
MELPA Stable is the recommended repo as it has the latest stable version. MELPA has a development snapshot for users who don't mind (infrequent) breakage but don't want to run from a git checkout.
You can install clojure-mode
using the following command:
M-x package-install [RET] clojure-mode [RET]
or if you'd rather keep it in your dotfiles:
(unless (package-installed-p 'clojure-mode)
(package-install 'clojure-mode))
If the installation doesn't work try refreshing the package list:
M-x package-refresh-contents
The clojure-mode
package actually bundles together several major modes:
clojure-mode
is a major mode for editing Clojure codeclojurescript-mode
is a major mode for editing ClojureScript codeclojurec-mode
is a major mode for editing.cljc
source files
All the major modes derive from clojure-mode
and provide more or less the same
functionality. Differences can be found mostly in the font-locking -
e.g. ClojureScript has some built-in constructs that are not present in Clojure.
The proper major mode is selected automatically based on the extension of the file you're editing.
Having separate major modes gives you the flexibility to attach different hooks
to them and to alter their behavior individually (e.g. add extra font-locking
just to clojurescript-mode
) .
Note that all modes derive from clojure-mode
, so things you add to
clojure-mode-hook
and clojure-mode-map
will affect all the derived modes as
well.
In the spirit of Emacs, pretty much everything you can think of in clojure-mode
is configurable.
To see a list of available configuration options do M-x customize-group RET clojure
.
The default indentation rules in clojure-mode
are derived from the
community Clojure Style Guide.
Please, refer to the guide for the general Clojure indentation rules.
By default multi-line docstrings are indented with 2 spaces, as this is a
somewhat common standard in the Clojure community. You can however adjust this
by modifying clojure-docstring-fill-prefix-width
. Set it to 0 if you don't
want multi-line docstrings to be indented at all (which is pretty common in most lisps).
The indentation of function forms is configured by the variable
clojure-indent-style
. It takes three possible values:
always-align
(the default)
(some-function
10
1
2)
(some-function 10
1
2)
always-indent
(some-function
10
1
2)
(some-function 10
1
2)
align-arguments
(some-function
10
1
2)
(some-function 10
1
2)
Note: Prior to clojure-mode 5.10 the configuration options for clojure-indent-style
used to be
keywords, but now they are symbols. Keywords will still be supported at least until clojure-mode 6.
The indentation of special forms and macros with bodies is controlled via
put-clojure-indent
, define-clojure-indent
and clojure-backtracking-indent
.
Nearly all special forms and built-in macros with bodies have special indentation
settings in clojure-mode
. You can add/alter the indentation settings in your
personal config. Let's assume you want to indent ->>
and ->
like this:
(->> something
ala
bala
portokala)
You can do so by putting the following in your config:
(put-clojure-indent '-> 1)
(put-clojure-indent '->> 1)
This means that the body of the ->/->>
is after the first argument.
A more compact way to do the same thing is:
(define-clojure-indent
(-> 1)
(->> 1))
You can also specify different indentation settings for symbols prefixed with some ns (or ns alias):
(put-clojure-indent 'do 0)
(put-clojure-indent 'my-ns/do 1)
The bodies of certain more complicated macros and special forms
(e.g. letfn
, deftype
, extend-protocol
, etc) are indented using
a contextual backtracking indentation method, require more sophisticated
indent specifications. Here are a few examples:
(define-clojure-indent
(implement '(1 (1)))
(letfn '(1 ((:defn)) nil))
(proxy '(2 nil nil (1)))
(reify '(:defn (1)))
(deftype '(2 nil nil (1)))
(defrecord '(2 nil nil (1)))
(specify '(1 (1)))
(specify '(1 (1))))
These follow the same rules as the :style/indent
metadata specified by cider-nrepl.
For instructions on how to write these specifications, see
this document.
The only difference is that you're allowed to use lists instead of vectors.
You can vertically align sexps with C-c SPC
. For instance, typing
this combo on the following form:
(def my-map
{:a-key 1
:other-key 2})
Leads to the following:
(def my-map
{:a-key 1
:other-key 2})
This can also be done automatically (as part of indentation) by
turning on clojure-align-forms-automatically
. This way it will
happen whenever you select some code and hit TAB
.
The available refactorings were originally created and maintained by the
clj-refactor.el
team. The ones implemented in Elisp only are gradually migrated
to clojure-mode
.
- Thread another expression.
Thread another form into the surrounding thread. Both ->>
and ->
variants
are supported. See demonstration on the
clj-refactor.el wiki.
- Unwind a threaded expression.
Supports both ->>
and ->
. See demonstration on the
clj-refactor.el wiki.
- Wrap in thread first (
->
) and fully thread.
Introduce the thread first macro and rewrite the entire form. With a prefix argument do not thread the last form. See demonstration on the clj-refactor.el wiki.
- Wrap in thread last (
->>
) and fully thread.
Introduce the thread last macro and rewrite the entire form. With a prefix argument do not thread the last form. See demonstration on the clj-refactor.el wiki.
- Fully unwind a threaded expression.
Unwind and remove the threading macro. See demonstration on the clj-refactor.el wiki.
- Cycle privacy
Cycle privacy of def
s or defn
s. Use metadata explicitly with setting
clojure-use-metadata-for-privacy
to t
for defn
s too. See demonstration
on the clj-refactor.el wiki.
- Cycle if/if-not
Find the closest if or if-not up the syntax tree and toggle it. Also transpose the "else" and "then" branches, keeping the semantics the same as before. See demonstration on the clj-refactor.el wiki.
Convert any given collection at point to list, quoted list, map, vector or set.
- Introduce let
Introduce a new let form. Put the current form into its binding form with a name provided by the user as a bound name. If called with a numeric prefix put the let form Nth level up in the form hierarchy. See demonstration on the clj-refactor.el wiki.
- Move to let
Move the current form to the closest let's binding form. Replace all occurrences of the form in the body of the let. See demonstration on the clj-refactor.el wiki.
- Forward slurp form into let
Slurp the next form after the let into the let. Replace all occurrences of the bound forms in the form added to the let form. If called with a prefix argument slurp the next n forms.
- Backward slurp form into let
Slurp the form before the let into the let. Replace all occurrences of the bound forms in the form added to the let form. If called with a prefix argument slurp the previous n forms.
paredit-convolute-sexp
is advised to replace occurrences of bound forms with their bound names when convolute is used on a let form.
- clojure-mode-extra-font-locking provides additional font-locking for built-in methods and macros. The font-locking is pretty imprecise, because it doesn't take namespaces into account and it won't font-lock a function at all possible positions in a sexp, but if you don't mind its imperfections you can easily enable it:
(require 'clojure-mode-extra-font-locking)
The code in clojure-mode-font-locking
used to be bundled with
clojure-mode
before version 3.0.
You can also use the code in this package as a basis for extending the font-locking further (e.g. functions/macros from more namespaces). Generally you should avoid adding special font-locking for things that don't have fairly unique names, as this will result in plenty of incorrect font-locking. CIDER users should avoid this package, as CIDER does its own dynamic font-locking, which is namespace-aware and doesn't produce almost any false positives.
-
clj-refactor provides refactoring support.
-
Enabling
CamelCase
support for editing commands(likeforward-word
,backward-word
, etc) inclojure-mode
is quite useful since we often have to deal with Java class and method names. The built-in Emacs minor modesubword-mode
provides such functionality:
(add-hook 'clojure-mode-hook #'subword-mode)
- The use of paredit when editing Clojure (or any other Lisp) code is highly recommended. It helps ensure the structure of your forms is not compromised and offers a number of operations that work on code structure at a higher level than just characters and words. To enable it for Clojure buffers:
(add-hook 'clojure-mode-hook #'paredit-mode)
- smartparens is an excellent
(newer) alternative to paredit. Many Clojure hackers have adopted it
recently and you might want to give it a try as well. To enable
smartparens
use the following code:
(add-hook 'clojure-mode-hook #'smartparens-strict-mode)
- RainbowDelimiters is a
minor mode which highlights parentheses, brackets, and braces
according to their depth. Each successive level is highlighted in a
different color. This makes it easy to spot matching delimiters,
orient yourself in the code, and tell which statements are at a
given depth. Assuming you've already installed
RainbowDelimiters
you can enable it like this:
(add-hook 'clojure-mode-hook #'rainbow-delimiters-mode)
- aggressive-indent-mode automatically adjust the indentation of your code,
while you're writing it. Using it together with
clojure-mode
is highly recommended. Provided you've already installedaggressive-indent-mode
you can enable it like this:
(add-hook 'clojure-mode-hook #'aggressive-indent-mode)
One of the fundamental aspects of Lisps in general, and Clojure in particular, is the notion of interactive programming - building your programs by continuously changing the state of the running Lisp program (as opposed to doing something more traditional like making a change and re-running the program afterwards to see the changes in action). To get the most of clojure-mode you'll have to combine it with some tool which will allow you to interact with your Clojure program (a.k.a. process/REPL).
A number of options exist for connecting to a running Clojure process and evaluating code interactively.
inf-clojure provides basic interaction with a Clojure REPL process.
It's very similar in nature and supported functionality to inferior-lisp-mode
for Common Lisp.
CIDER is a powerful Clojure interactive development environment, similar to SLIME for Common Lisp.
If you're into Clojure and Emacs you should definitely check it out.
An extensive changelog is available here.
Copyright © 2007-2018 Jeffrey Chu, Lennart Staflin, Phil Hagelberg, Bozhidar Batsov, Artur Malabarba and contributors.
Distributed under the GNU General Public License; type C-h C-c to view it.